Data Driven Guides:Supporting expressive design for Information graphics
Nam Wook KimHarvard
Zhicheng Leo LiuAdobe
Mira DontchevaAdobe
Hanspeter PfisterHarvard
Wilmot LiAdobe
Jovan Popović Adobe
Eston SchweickartCornell
Source: Google Search Trends2016
Analysts &Researchers
Source: Google Search Trends
Artists, Journalists,
Bloggers, Designers
Source: Google Search Trends
Visualization Design Tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
Visualization Design Tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Less expressive(Automatic)
More expressive(Manual)
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Plot.lyExcel
Spotfire
Tableau
ManyEyes
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Visualization Design Tools
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Visualization Design Tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Less expressive(Automatic)
More expressive(Manual)
Visualization Design Tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
?
Interviews with infographic designers.
Interviews with infographic designers.Participants • 2 professional designers
3 master student designers1 visualization researcher.
Interviews with infographic designers.Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
Interviews with infographic designers.Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
• Mostly use vector editing tools such as Adobe Illustrator.
Interviews with infographic designers.Questions • What difficulties they face in
creating infographics?
Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
• Mostly use vector editing tools such as Adobe Illustrator.
Interviews with infographic designers.Questions • What difficulties they face in
creating infographics?
• Overall design practice.
Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
• Mostly use vector editing tools such as Adobe Illustrator.
Interviews with infographic designers.Questions • What difficulties they face in
creating infographics?
• Overall design practice.
• How they manually encode data into graphics.
Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
• Mostly use vector editing tools such as Adobe Illustrator.
Interviews with infographic designers.Questions • What difficulties they face in
creating infographics?
• Overall design practice.
• How they manually encode data into graphics.
Related work Bigelow et al (AVI, 2014).
Participants • 2 professional designers
3 master student designers1 visualization researcher.
• 2 ~ 10 years of experience in graphic & infographic design.
• Mostly use vector editing tools such as Adobe Illustrator.
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
1. Lack of flexible design in visualization construction tools
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
Difficult to add annotations & embellishments.
`
1. Lack of flexible design in visualization construction tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
Difficult to design new visual marks & layouts.
`
`
Difficult to add annotations & embellishments.
1. Lack of flexible design in visualization construction tools
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Less expressive(Automatic)
More expressive(Manual)
2. Tedious manual encoding required in graphic design tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)
2. Tedious manual encoding required in graphic design tools
Custom Scale
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Less expressive(Automatic)
More expressive(Manual)
3. Absence of data binding for custom/imported charts in graphic design tools
3. Absence of data binding for custom/imported charts in graphic design tools
IllustratorCorelDRAWAffinityDesigner
InDesignInkscape Photoshop
Plot.lyExcel
Spotfire
Tableau
ManyEyes
LyraiVisDesigner
SageBrushDataVisual
Improvise
Less expressive(Automatic)
More expressive(Manual)Graph Tool in Illustrator.
To customize this chart, ungrouping is required resulting in loss of data binding.
Design Goals
1. Maintain flexibility in the design process.
e.g., not enforcing a predefined outcome or specific order of operations.
Design Goals
2. Provide methods for accurate data-driven drawing.
1. Maintain flexibility in the design process.
Design Goals
Design Goals
2. Provide methods for accurate data-driven drawing.
1. Maintain flexibility in the design process.
3. Support persistent data binding for freeform graphics.
Data Driven Guides:Supporting expressive design for Information graphics
Length guide
a
Area guide
d = aread = length
Data Table
A B C D
0
18
35
53
70
Graph
Data Table
A B C D
0
18
35
53
70
Graph
Data-Driven GuidesData Table
A B C D
Flexible direct manipulationTo manipulate to create a custom layout
d2
d1d3
d4
Maintain proportional lengths To preserve data integrity
Snap to a guide To support accurate data-driven drawing
Data-binding using guidesas the backbone of associated shapes.
(2D deformation for vector graphic)
Architectural Design Drawing
Logo Design User Interface Design
Designer’ Friends: Rulers, Grids, Guides
Length and area guides can also be used as a Position guide.
Length guide
a
Area guide
d = length d = area
Default guide color: Cyan
J. Bertin.,1983.Rankings of visual variables by J. Mackinlay.,1986.
Fundamental channels for encoding information .Visual variables
J. Bertin.,1983.Rankings of visual variables by J. Mackinlay.,1986.
Fundamental channels for encoding information .
Frequently used in infographics, although
often misused.
Visual variables
Use Cases
Use cases1. Drawing data-driven graphics
Cyan: Data-Driven GuidesCyan: Data-Driven Guides
Cyan: Data-Driven Guides
Create & manipulate
Link & Repeat
Radial layout & link inspection
Composite structure & copy and paste
Label generation
Deforms
Linear blend skinning
Related Work 1. Bounded biharmonic weights for real-time deformation. Jacobson, Alec, et al. ACM Trans. Graph., 2011 2. Skinning cubic Bézier splines and Catmull-Clark subdivision surfaces. Liu, Songrun, et al. ACM Trans. Graph., 2014.
Deforms
Linear blend skinning
Old point in a shape.
Related Work 1. Bounded biharmonic weights for real-time deformation. Jacobson, Alec, et al. ACM Trans. Graph., 2011 2. Skinning cubic Bézier splines and Catmull-Clark subdivision surfaces. Liu, Songrun, et al. ACM Trans. Graph., 2014.
Rest pose
Deformed
Rest pose
Deformed Linear blend skinning
Related Work 1. Bounded biharmonic weights for real-time deformation. Jacobson, Alec, et al. ACM Trans. Graph., 2011 2. Skinning cubic Bézier splines and Catmull-Clark subdivision surfaces. Liu, Songrun, et al. ACM Trans. Graph., 2014.
Spatial Transformations(changes in guides)
Linear blend skinning
Related Work 1. Bounded biharmonic weights for real-time deformation. Jacobson, Alec, et al. ACM Trans. Graph., 2011 2. Skinning cubic Bézier splines and Catmull-Clark subdivision surfaces. Liu, Songrun, et al. ACM Trans. Graph., 2014.
Skinning weight(Transformation amount)
Rest pose
Deformed
Linear blend skinning
Related Work 1. Bounded biharmonic weights for real-time deformation. Jacobson, Alec, et al. ACM Trans. Graph., 2011 2. Skinning cubic Bézier splines and Catmull-Clark subdivision surfaces. Liu, Songrun, et al. ACM Trans. Graph., 2014.
New points in a shape.
Rest pose
Deformed
Deforms
MovesMoves
Used as a Position Guide
Used as flexible rulers.
Cyan: Data-Driven Guides
Cyan: Data-Driven Guides
Four area guides are usedto encode a single shape.
Use cases1. Drawing data-driven graphics
2. Retargeting existing artworks
Original By Nigel Holmes
Original By Nigel Holmes
Use cases1. Drawing data-driven graphics
2. Retargeting existing artworks
3. Proofreading existing infographics
Proofreading existing infographics
By Nigel Holmes. By Tiffany Farrant-Gonzalez.
Cyan: Data-Driven Guides
Proofreading existing infographics
By Nigel Holmes.
Original
The length of marks do not match the length of guides (curved).
Proofreading existing infographics
By Tiffany Farrant-Gonzalez.
Area guide
Length guide (Radius)
Original
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
Informal usability evaluation with designers.
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
• 2 ~ 10 years of experience in infographic design.
Informal usability evaluation with designers.
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
• 2 ~ 10 years of experience in infographic design.
• Frequently used tools include vector & image editors(programming: only 2 people).
Informal usability evaluation with designers.
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
• 2 ~ 10 years of experience in infographic design.
• Frequently used tools include vector & image editors(programming: only 2 people).
Informal usability evaluation with designers.Procedure • A 60 min session with a 15 min
tutorial
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
• 2 ~ 10 years of experience in infographic design.
• Frequently used tools include vector & image editors(programming: only 2 people).
Informal usability evaluation with designers.Procedure • A 60 min session with a 15 min
tutorial
• Pre-task and post-task surveys
Participants • 13 master student designers
(architecture, urban planning, and infographic design etc)
• 2 ~ 10 years of experience in infographic design.
• Frequently used tools include vector & image editors(programming: only 2 people).
Informal usability evaluation with designers.Procedure • A 60 min session with a 15 min
tutorial
• Pre-task and post-task surveys
• 2 replication tasks
• 1 creative task
* We did not measure time
Data: GDP of G5 Countries.
Creative Tasks • Two selected infographics
created by participants.
Results
Creative Tasks • Two selected infographics
created by participants.
ResultsPost-task surveys (5-point Likert scale) 1. Interactions with DDG were intuitive.
(μ=4.0, σ=0.71)
Data: GDP of G5 Countries.
Creative Tasks • Two selected infographics
created by participants.
ResultsPost-task surveys (5-point Likert scale) 1. Interactions with DDG were intuitive.
(μ=4.0, σ=0.71)
2. DDG is useful for positioning and measuring custom shapes based on data compared to rulers or grids.(μ=4.7, σ=0.63)
Data: GDP of G5 Countries.
Creative Tasks • Two selected infographics
created by participants.
ResultsPost-task surveys (5-point Likert scale) 1. Interactions with DDG were intuitive.
(μ=4.0, σ=0.71)
2. DDG is useful for positioning and measuring custom shapes based on data compared to rulers or grids.(μ=4.7, σ=0.63)
3. DDG is useful for designing creative and expressive infographics(μ=4.9, σ=0.38)Data: GDP of G5 Countries.
“Currently, I need a calculator to make data graphic, which is pretty arduous. This tool makes it much easier to try things out and experiment with the graphics.” - P10.
Results: qualitative feedback
“Currently, I need a calculator to make data graphic, which is pretty arduous. This tool makes it much easier to try things out and experiment with the graphics.” - P10.
“I think that this would be a wonderful aid in creating graphics for architectural representations.” - P3.
Results: qualitative feedback
“Currently, I need a calculator to make data graphic, which is pretty arduous. This tool makes it much easier to try things out and experiment with the graphics.” - P10.
“I think that this would be a wonderful aid in creating graphics for architectural representations.” - P3.
“I usually do very analytical infographics, using traditional forms like bars or circles. Because of that, I’m not quite sure if data guides might be very useful” - P5.
Results: qualitative feedback
1. Data-driven guides currently works with a tabular dataset.
Limitations
1. Data-driven guides currently works with a tabular dataset.
2. Difficult to generate guide elements such as axes and legends.
Limitations
1. Data-driven guides currently works with a tabular dataset.
2. Difficult to generate guide elements such as axes and legends.
3. Other visual variables have to be manually encoded such color or angle.
Limitations
Template
Future work1. Reusable creative infographic templates.
Template
Future work1. Reusable creative infographic templates.
2. Data-driven guides for other visual variables.
Color, angle, shape, slope etc.
Template
Future work1. Reusable creative infographic templates.
2. Data-driven guides for other visual variables.
Color, angle, shape, slope etc.
3. Intelligent systems for automatically providing design feedback.
Area vs radiusIncorrect size
Data Driven Guides:Supporting expressive design for Information graphics
www.namwkim.org/ddg